Amy Carpenter Aquino
January 2023—For the central microbiology laboratory serving Barnes-Jewish and four other hospitals in the St. Louis area, validating and implementing a new blood culture system and moving to a shorter incubation time came at a perfect time: right before the pandemic.
“We implemented this just a few months before we saw record-breaking hospitalizations at our institution,” Eric M. Ransom, PhD, D(ABMM), said of the work done on the BacT/Alert Virtuo blood culture detection system (BioMérieux) at Washington University in St. Louis in 2019. Dr. Ransom, now assistant medical director of clinical microbiology at University Hospitals Cleveland Medical Center, reported at the AACC meeting last July on the Virtuo implementation, including a study finding that a four-day incubation time was sufficient. “We wouldn’t have had the space on our blood culture instruments if we didn’t have all that free space we had generated by increasing our capacity 20 percent,” he said (Ransom EM, et al. J Clin Microbiol. 2021;59[3]:e02459-20).
Dr. Ransom, who is also assistant professor of pathology at Case Western Reserve University School of Medicine, said the pandemic forced some institutions to go to a four-day incubation, “so this ended up being important information to get out there in the literature.”
Workflow improvements were immediately apparent in the laboratory’s switch from the VersaTrek (Thermo Scientific) to the Virtuo system, which Dr. Ransom described as composed of an A unit and three lookalike B units. The BacT/Alert resin culture bottles used with Virtuo are more durable in transit, he said, than the glass bottles used previously. The blood culture set—one FA Plus (aerobic) bottle and one FN Plus (anaerobic) bottle—is loaded onto the conveyer belt on the A unit, after which staff can walk away.
“You’re not opening the door, releasing all of that nice, incubated warm air at the perfect 35, 37 degrees,” Dr. Ransom said. Virtuo is “a mostly closed system in which a very small door opens, maintaining that temperature.” Positive bottles drop out with no need to open the door for retrieval, he said, noting that an alarm sounds to alert the staff because it’s a stat test. Negative bottles drop from the machine into a discard bin at the bottom, again eliminating the need to open the door.
Dr. Ransom recommends the Clinical and Laboratory Standards Institute’s guideline “M47: Principles and Procedures for Blood Cultures,” second edition, published in 2022. Its authors advise, when implementing a new blood culture system, testing the new and existing systems in parallel (by collecting two culture sets with equal blood volumes) and comparing them head to head. They recommend obtaining at least 20 positives, “ideally real-world pathogens” and not contaminants, Dr. Ransom said. “And 95 percent of these results should agree with the current system to be considered equivalent.” The low positivity rate of blood cultures—five to 12 percent at most institutions—makes this good approach difficult, he said. “So you’re going to have to test a lot more blood cultures to get those 20.” A high contamination rate adds to the difficulty.
Whether it’s ethical to divert patient blood to validate the new system is a question, he said. “We sure hope our new system is going to work equally well or better, but in a situation where it may not, you’re taking all that blood from the patient and putting it in a new blood culture system” when it could have been used for the existing system.
A seeded blood culture study is an alternative or can be done in parallel. “Here you have a lot more control—which organisms you’re going to test, making sure they represent the common pathogens at an individual facility and are pertinent to those patient populations.” And “here you can control that starting inoculum, which would be important when looking at times to positivity.” He recommends a common range of five to 10 CFUs per bottle. (For detail on dilutions and more, he suggests Clinical Microbiology Procedures Handbook, fifth edition.)
[dropcap]W[/dropcap]hether performing the clinical side-by-side comparison or the seeded study, Dr. Ransom advises a close look at organism recovery to ensure the same organisms are recovered at the same rate on both systems. “Compare your time to positivity,” he said. “Hopefully it’s equal to if not better than your current system.” And evaluate all media.
He advises the following also:
- Establish the acceptance criteria early on to avoid the temptation of later adjustments to ensure passing results. “Just because something doesn’t reach that threshold doesn’t mean it fails, but it may warrant additional investigation or review,” he said.
- Account for all parts of the blood culture workflow, including bottle processing, which may require workforce retraining. “A lot of these bottles have unique adapters to get that blood out for subsequent testing, like Gram stains,” he said. Since some Gram stain differences may be slight, “you may need to adjust your times for different steps in new Gram stains to make sure you’re getting equivalent results if you’re using a new blood culture media.”
- Account for different thresholds of detecting organisms on blood culture systems as rapid diagnostic methods or direct aspartate aminotransferase testing on positive blood culture broth become more popular, to avoid “a negative result downstream,” Dr. Ransom said. And keep in mind that sterile body fluids or platelets or other non-blood specimens may be in blood culture bottles and need to be evaluated also.
At Barnes-Jewish Hospital, a combination of clinical and seeded study methods was used to validate the Virtuo. The seeded study used aerobic, anaerobic, and fastidious organisms and yeast, and each was tested in triplicate for each bottle type. “We really kicked the tires on this validation, largely because at the time Virtuo was very new to the market, there wasn’t a lot of published literature on it. We wanted to make sure it was robust and held to our standard,” Dr. Ransom said. They looked at delayed entry onto the machine, compared instrument bank one to instrument bank two, assessed sterile body fluids and platelet products, and used the Verigene Gram-positive assay for the new blood culture media assessment.
“Overall, it was very comparable, very equivalent,” Dr. Ransom said. “As far as organism recovery, we were able to recover the organisms, regardless of the system.” The study authors saw a difference in time to positivity among organisms. “Thankfully, that was in our favor. The new system ended up having a shorter time to positivity, allowing us to get those results out quicker to our providers.” In the subset of organisms tested, the most significant time difference was found in Staphylococcus aureus, which had a mean time to positivity of 10.6 hours on Virtuo versus 12.4 hours on VersaTrek.
“Unfortunately, not all organisms saw improved growth rates,” he said. Clostridium septicum, for example, had an average time to positivity of 22.1 hours on Virtuo versus an average TTP of 12.8 hours on VersaTrek. For Candida albicans, it was 26.8 hours on Virtuo and 23.6 on VersaTrek. And for Cryptococcus neoformans, 67.8 (Virtuo) and 64.2 (VersaTrek).
The validation went well, and rollout was expected. “But then we started doing more assessments,” Dr. Ransom said, one of which was of the delay in transport times from hospitals at a distance from the centralized microbiology laboratory.
When the microbiology laboratory evaluated time to positivity after a delayed entry on the Virtuo, which involved inoculating bottles and allowing one bottle to sit at room temperature while the other was put directly on the machine, most bacteria did not show a significant difference, he said. Streptococcus pneumoniae showed an average difference of -0.9 hours in TTP in the FA Plus (aerobic) bottles and a difference of -0.4 hours in TTP in the FN Plus (anaerobic) bottles after a delayed entry of eight hours. E. coli was actually “faster” with delayed entry. “But that’s merely showing the microbe itself being able to replicate fairly quickly at room temperature, so it had a head start.”
Neisseria gonorrhoeae had a “bit of a delay” in TTP (average difference of 7.1 hours in aerobic bottles). Bacteroides fragilis “only grew anaerobically and it was very comparable across,” Dr. Ransom said. “This was great, reassuring us that we could maintain being a centralized laboratory, having those specimens coming in even if they wouldn’t get on the machine for eight hours.”
[dropcap]A[/dropcap] study of the impact of the Virtuo system implementation at Barnes-Jewish Hospital was published in 2021 (Chavez MA, et al. J Clin Microbiol. 2021;59[10]:e0061721). The authors retrospectively reviewed all blood cultures performed in 2018 (VersaTrek) and in 2019 (Virtuo) and compared positivity and contamination rates. The positivity rate increased from 8.1 percent before Virtuo was implemented to 11.7 percent post-implementation. In its investigation, the laboratory found a spike post-implementation in what it considered to be contaminants, Dr. Ransom said. Its first step was to reeducate to ensure the new bottles were collected properly.
Driving much of the increase, the data showed, was a higher positivity rate of the staphylococci. S. aureus positivity, in particular, increased from 1.5 percent of blood cultures pre-implementation to 3.4 percent of blood cultures post-implementation. “It turns out that a big contributor to this was that we were detecting bacteria in the bloodstream longer than we were with our older system,” Dr. Ransom said. The improved beads in the new BacT/Alert bottles were “probably inhibiting those antimicrobials a little better.”
Pre-implementation, a blood culture would be sent, and if it tested positive, the patient would be treated. “You may detect it that first time, maybe a day later,” he said. Blood cultures were drawn maybe two days later. “But we were seeing a shift, and overall we were detecting it on day three and day four,” he said, which worried providers initially. “A lot of education had to happen,” Dr. Ransom said. But there was no real impact on patient health.
The microbiology team investigated also whether a five-day incubation was still appropriate using the Virtuo system, an investigation prompted by what Dr. Ransom heard from technologists during rounds. At least once a week, he would hear that a culture went positive on day four. “It’s going to be mixed. Why are we working up these cultures?” they would ask. He would reassure them the cultures were worth doing, and they would say that once they relayed the result to the provider, the provider would say, “The patient’s already been discharged. This person’s fine.” Hearing that time and time again, Dr. Ransom thought “Maybe they’re onto something.”
As incubation time increases, he noted, so does the chance of detecting all true bloodstream infections. “Unfortunately, this is not perfect,” he said, because a longer incubation also increases the chances of contamination growth. “It’s a fine balance in figuring out where you want to have this cutoff.”
A seven-day incubation time was more common in the ‘90s and 2000s, he said, noting the transition to five days was largely owing to better instrumentation and media. Was it time to look at reducing the incubation interval again with the Virtuo, which held temperature at a more constant rate and had improved media?
In their study, Dr. Ransom and colleagues looked at all blood cultures that came through the central microbiology laboratory (from fall 2018 to fall 2019) from the five hospitals. They examined the outcomes of all positive bottles and time to positivity, defining TTP as the time from when the bottles were loaded onto the instrument to when they came off and signaled positive. Medical charts were reviewed for patients with a bottle with TTP greater than four days.
In studying the TTP of 13,592 positive bottles during the study period, Dr. Ransom and colleagues found 76.74 percent of all positives (10,431 bottles) were detected within the first 24 hours. After day two, the positivity rate was 93.56 percent; after day three, 97.07 percent, and after day four, 98.71 percent, “leaving just 175 bottles in those final 24 hours,” he said.
They compared adult TTP data (12,769 bottles) with pediatric data (823 bottles) and the two data sets “matched almost perfectly,” he said. “If anything, it [TTP] was slightly faster” in pediatric blood cultures: 76.55 percent positive in the first 24 hours, 95.75 percent positive at end of day two, 98.42 percent positive at day three, and 99.39 percent positive at day four, leaving only five bottles becoming positive in the final 24 hours.
S. aureus (3,893 bottles) and S. epidermidis (1,248) grew quickly, and E. coli (1,382) “was one of our most impressive organisms,” with 94.3 percent found to be positive in the first 24 hours, the vast majority positive within 16 hours, and 50 percent positive in 10 hours, he said. S. pneumoniae (192) grew exceptionally well: 98.4 percent at day one, 100 percent at day two. Pseudomonas aeruginosa (298) took a little longer than some of the other organisms, “but still at the end of day three, 98 percent; by the end of day four, 99.3 percent.”
With so few positives in the final 24 hours, they asked, do they need them? If those bottles are driving treatment and saving lives, he said, “we want to maintain them or potentially extend our incubation time.”
A medical chart review of those 175 bottles (0.1 percent of all bottles and 1.3 percent of positive bottles) revealed 160 had no impact at all, with 38 noted in the record as contaminated—not a surprising number, he said, given the cultures were from four days prior.
The other 15 bottles had impact, but in some cases the impact was negative—additional cultures or further workup later deemed unnecessary or a delay in discharge, for example. “We also had examples in which the patient had already been discharged, and the provider, then hearing there’s a positive, had that patient go to the ED or to their primary care physician,” he said. By the most conservative estimate, fewer than 10 bottles were clinically actionable.
The study findings resulted in their implementing a four-day incubation protocol. “Initially there was a lot of hesitancy,” but once the data were made known, physicians asked, “How quickly can we make this happen?” Dr. Ransom said. They were given the opportunity to request a five-day incubation period. “Nobody calls the laboratory asking for that additional 24 hours back or longer,” he said.
Since 90 percent of blood cultures are negative, the vast majority of the bottles received in the laboratory are going to become final as negative one day earlier, Dr. Ransom said. “That’s earlier discharges for patients,” and earlier deescalation of treatment—a plus for the patient, physician, and laboratory. Another benefit: The recovery rate of the contaminant Cutibacterium spp. declined from 0.87 percent to 0.23 percent. If the contamination rate or positivity rate is increased, “you also increase chances of a central-line associated bloodstream infection,” and many eyes in an institution are on that number, he said, because such infections can affect reimbursement. “This is something the hospital looks at very closely and you want to be sure you’re investigating thoroughly.”
Amy Carpenter Aquino is CAP TODAY senior editor. Carey-Ann Burnham, PhD, D(ABMM), Melanie Yarbrough, PhD, D(ABMM), D(ABCC), and the laboratory staff at Barnes-Jewish Hospital were central to the studies reported.